公开(公告)号：KR1020090001027A

公开(公告)日：2009-01-08

申请号：KR1020070065083

申请日：2007-06-29

Applicant: 조선대학교산학협력단

Inventor： 김판구 ,  황명권 ,  백선경 ,  황광수

IPC: G06Q50/10 ,  H04L9/32 ,  H04N21/845 ,  G06F17/30

CPC classification number: G06Q50/10 ,  G06F17/30784 ,  H04L9/32 ,  H04N21/845

Abstract: A web based image reuse system and a method are provided to reduce that the cost in the manufacture and the labor and time by reusing the photography and the images which are already manufactured, while a user directly does not take a picture of necessary images in a moving picture production. A web based image reuse system comprises: a user authentication server(310) for controlling the access of a user through an authentication of a user access apparatus(210); a query processing server(320) for matching the images which coincide with a multimedia server(330) by analyzing a query language delivered through a connecting apparatus; and a multimedia server for storing images uploaded from a plurality of users through the connecting apparatus.

Abstract translation: 提供了一种基于网络的图像重用系统和方法，用于通过重复使用已经制造的图像和图像来减少制造和劳动和时间的成本，而用户直接不在图像中拍摄必要的图像 活动图片制作。 基于网络的图像重用系统包括：用户认证服务器（310），用于通过用户访问设备（210）的认证来控制用户的访问; 查询处理服务器（320），用于通过分析通过连接装置传递的查询语言来匹配与多媒体服务器（330）重合的图像; 以及多媒体服务器，用于存储通过连接装置从多个用户上传的图像。

公开(公告)号：KR100820746B1

公开(公告)日：2008-04-11

申请号：KR1020070006558

申请日：2007-01-22

Applicant: 조선대학교산학협력단

Inventor： 김판구 ,  공현장 ,  황명권

IPC: G06F17/30

CPC classification number: G06F17/30734 ,  G06F17/30675 ,  G06F17/30716 ,  G06F17/30864

Abstract: A system and a method for browsing information based on ontology are provided to offer information efficiently up to attribute concepts used for defining a keyword, and upper, lower, synonym, and anonym concepts related to the keyword by browsing the searched information based on an ontology graph model. An ontology server(310) includes definitions for tangible/intangible concepts by using ontology description language. A browsing service server(340) includes an ontology analyzer(341) extracting a triple by analyzing ontology of the ontology server, an ontology inferrer(342) inferring a new fact based on the analyzed tripe, a query matcher(344) extracting the concepts related to a query by matching the query received from a user with the ontology, a graph structure representer(343) moving the concepts and reforming a graph based on the query by converting the extracted triples into the graph, and a search result processor(345) moving and reforming the graph when the user selects the concept in the graph. A web information analysis server(330) includes a web document analyzer(331) and a new concept/attribute processor(332). An ontology management server(320) includes an ontology updater(322) and an integrated ontology manager(321).

Abstract translation: 提供一种用于浏览基于本体的信息的系统和方法，以便有效地提供信息，以便通过基于本体论浏览搜索到的信息来定义用于定义关键字的概念，以及与关键字相关的上，下，同义词和匿名概念 图形模型。 本体服务器（310）通过使用本体描述语言包括有形/无形概念的定义。 浏览服务服务器（340）包括本体分析器（341），通过分析本体服务器的本体来提取三元组，本体推断器（342）基于所分析的分析来推断新事实，提取​​概念的查询匹配器（344） 通过将从用户接收的查询与本体相匹配来查询查询，图形结构表示器（343），通过将提取的三元组转换为图形，基于查询来移动概念和重建图形;以及搜索结果处理器（345 ）当用户选择图中的概念时，移动和重构图形。 网络信息分析服务器（330）包括网络文档分析器（331）和新的概念/属性处理器（332）。 本体管理服务器（320）包括本体更新器（322）和集成本体管理器（321）。

公开(公告)号：KR1020080087920A

公开(公告)日：2008-10-02

申请号：KR1020070007610

申请日：2007-01-24

Applicant: 조선대학교산학협력단

Inventor： 김판구 ,  황명권

IPC: G06Q50/10 ,  G06F17/30

CPC classification number: G06Q50/10 ,  G06F17/30017

Abstract: A method for comparing an object with multimedia information to apply sensibility to multimedia retrieval is provided to apply to various kinds of designs considering the sensibility, and realize effective utilization for various types of multimedia retrievals and production requested from a user. A slope of an object is found by drawing a horizontal line(500) at a point closed first at a lower end of the object. 2D(Dimensional) tangent space representation(300) is found by finding an angle of each vertex and a length between the vertexes from the furthest left point(400) in an anticlockwise direction. A difference between two sensibilities is represented as much as areas(310-340) compared by using the tangent space representations.

Abstract translation: 提供了一种用于将对象与多媒体信息进行比较以将感性应用于多媒体检索的方法，以应用于感觉的各种设计，并且实现从用户请求的各种类型的多媒体检索和生产的有效利用。 通过在对象的下端首先关闭的点绘制水平线（500）来找到对象的斜率。 通过在逆时针方向上从最左边的点（400）找到每个顶点的角度和顶点之间的长度来找到2D（尺寸）切线空间表示（300）。 通过使用切线空间表示比较区域（310-340），表示两个感觉之间的差异。

公开(公告)号：KR1020080087063A

公开(公告)日：2008-09-30

申请号：KR1020070007635

申请日：2007-01-24

Applicant: 조선대학교산학협력단

Inventor： 김판구 ,  황명권 ,  황광수

IPC: G06F17/27 ,  G06F17/30

CPC classification number: G06F17/2785 ,  G06F17/2705 ,  G06F17/30011

Abstract: A document comparing method using document hierarchy is provided to represent contents of documents as significant layers and give an important domain a weight to achieve significant document search. Documents are matched to a word net to extract nouns(110), and SYNSET ID is detected using adapted LESK algorithm(120). A hierarchy structure is grasped through the detected SYNSET ID(130), and the hierarchy structure is converted into an RDF(Resource Description Framework) document(140). A domain constructing each layer of the hierarchy structure is detected, and the weight of each domain in the hierarchy is checked(220). Domains detected for respective documents are compared(210).

Abstract translation: 提供了使用文档层次结构的文档比较方法来将文档的内容表示为重要的层次，并给出重要的域以实现重要的文档搜索。 文档与单词网匹配以提取名词（110），并且使用适配的LESK算法（120）检测SYNSET ID。 通过所检测的SYNSET ID（130）来掌握层次结构，并且将层次结构转换为RDF（资源描述框架）文档（140）。 检测构造层次结构的每个层的域，并且检查层次结构中每个域的权重（220）。 对各文档检测到的域进行比较（210）。

公开(公告)号：KR1020090048812A

公开(公告)日：2009-05-15

申请号：KR1020070114863

申请日：2007-11-12

Applicant: 조선대학교산학협력단

Inventor： 김판구 ,  황명권 ,  황광수 ,  백선경

IPC: G06F17/30

CPC classification number: G06F17/30268

Abstract: 사용자가 입력한 질의어와 가장 유사한 이미지를 찾아주기 위하여 이미지 데이터에서 키워드를 추출하고, 워드넷을 이용하여 키워드간 유사도를 측정하며, 측정한 유사도를 이용하여 이미지 우선순위를 분석하여 의미가 질의어와 가장 가까운 이미지를 찾아주는 방법
유사도, 이미지 검색, 의미량, 워드넷

公开(公告)号：KR1020090048261A

公开(公告)日：2009-05-13

申请号：KR1020070114579

申请日：2007-11-09

Applicant: 조선대학교산학협력단

Inventor： 김판구 ,  황명권 ,  황광수 ,  백선경

IPC: G06F17/40

CPC classification number: G06F17/30705 ,  G06F17/2735 ,  G06F17/2785 ,  G06F17/30616

Abstract: 국내문서의 효율적이고 의미적인 분류를 위하여 의미사전을 이용하여 문서내 명사를 추출하며, 추출한 개념의 빈도값(TF: Term Frequency)과 주변 개념들과의 관계된 정도(RV: Relation Value)를 측정하고, 추출된 키워드들을 의미사전과 매핑하여 핵심 키워드 그룹 파악하고, 개념가중치를 구하여 의미사전에 매핑하는 과정을 통하여 의미적인 문서 분류 방법
주제선정, 핵심키워드그룹, 자동분류, 의미사전